Like big data, analytics is a major topic these days. But what does this buzzword refer to — and how do companies stand to gain? We provide an overview of the use of analytics in customer communication and explain the terms descriptive, predictive and prescriptive analytics.
What does “analytics” really mean? The word analysis derives from Greek and means the breaking down of something complex into its individual parts and then examining them. In customer service, analytics is understood to mean the analysis of large volumes of data in order to arrive at insights regarding customer behavior. Various levels of analysis can be distinguished: descriptive analytics, predictive analytics and prescriptive analytics.
Descriptive analytics is fundamentally “classic business intelligence”: historical data is grouped and portrayed according to different dimensions in order to find out where problems or untapped potential lie. With this, a company can tell, for example, which customers are active or shop frequently, or which ones purchase a large amount during each visit. Based on this information, advertising flyers and emails can be managed. For example, these campaigns can include personalized discount coupons as incentives to purchase more. Inactive customers could receive credit toward their next purchase, and less frequent customers could be given a discount after they purchase a certain amount. “But genuine individualization of advertising and CRM measures on the basis of descriptive analytics doesn’t normally happen,” explains Dirk Schäfer, Director of CRM Analytics at Arvato CRM Solutions.
While descriptive analytics works with grouped historical data, predictive analytics uses this data to generate additional knowledge: in this case, the idea is to “look into the future” and create forecasts to describe, for instance, how likely it is that a particular customer will buy a certain product or redeem a certain coupon. The basis for this is person-specific historical data: core data, order history, click streams, etc. A campaign that uses predictive analytics might offer customers precisely the products that are best suited to their current personal buying interests. Another field of application is predicting customer behavior: what might the customer be interested in at the next service contact; will the customer cancel soon, and how certain is it that the next order will be paid. “With predictive analytics, companies have the opportunity to identify and predict the customer’s interests and activities. They can therefore address each customer individually and in a targeted fashion,” says Dirk Schäfer.
Both analytics methods thus create customer knowledge that can be used in marketing and customer service for promising, customer-oriented measures. With the aid of analytical results, useful decisions can be made. But aren’t there more and more important touchpoints all the time in modern, online business relations? And aren’t customers reacting ever faster, and in ways that are increasingly context-dependent? And as a result, don’t decisions have to be made immediately and automatically? This is where prescriptive analytics comes into play. Here, the existing historical data as well as the current situation are taken into account in order to determine precisely the right product or offer for a particular customer at this moment on this channel. For example, this might be when the customer has just browsed the online shop or phoned a customer manager. “Prescriptive analytics answers the question: what should I suggest to the customer in order to improve a certain strategic company performance indicator,” explains Schäfer. A recommended action (NBA: “next best action”) could be a personalized discount, which is determined in real time during a purchase. “With this service, which we already offer our partners, companies can use data in real time to optimize their dialogue with customers.”
The future of customer interaction
In light of the ever-increasing number of touchpoints and the growing customer demands, analytics will play a central role in interactions with customers and the service they receive. Its importance will actually even grow as more use is made of the huge amount of available data from Internet-connected devices such as smartphones, intelligent home appliances or connected cars. Driving behavior could be analyzed, for example, in order to give the driver tips for saving fuel and reducing wear on the tires. Sensors can record all the vehicle’s technical and service-related data and give appropriate recommendations, such as a visit to the garage. Of course, the appointment can be arranged directly via the car — either through the hotline or by the chatbot in the car, which could coordinate the visit with the service partner’s bot. When the fuel tank is nearly empty or the battery depleted, the car’s monitor will reveal where and at what price more fuel can be bought, or where the battery can be recharged. And even when drivers are not sitting in the car, they are connected to it and can see on their smartphone the way to the car parking lot and the remaining parking time. This is only one example of the customer interaction of the future with its highly personalized and individual real-time services.